Intel BX80619I73820, Design Manual

Page: 78 / 124

PECI Interface
78 Thermal/Mechanical Specifications and Design Guide
7.1.4.6 Socket Power Limit Performance Status Read
This service allows the PECI host to assess the performance impact of the currently
active power limiting modes. The read return data contains the total amount of time for
which the entire processor package has been operating in a power state that is lower
than what the operating system originally requested. This information is tracked by a
32-bit counter that wraps around. The unit for time is determined as per the Package
Power SKU Unit settings described in
Section 7.1.2.7.2 .
7.1.4.7 Socket Power-Performance Scalability Read
This feature allows the PECI host to access time-averaged processor package power-
performance ratio information at the current workload and operating point. This is
typically used to maximize overall platform performance-power ratios by appropriate
budgeting between multiple packages.
7.1.4.8 Socket Performance Indication Read
This read returns information on the total number of retired instructions and is
monitored by a 32-bit counter. It may be used in conjunction with the ‘Accumulated
Run Time Read’ described in
Section 7.1.2.7.5 to track the number of retired
instructions per time unit. This feature is typically used for power budgeting through
appropriate distribution of power across racks of heterogeneous platforms. Note the
following equation for calculation for the actual number of retried instructions.
Number of retried instructions = Socket Performance Indication Data * 128
Figure 7-34. Socket Power Limit Performance Data
Accumulated CPU Throttle Time
Accumulated CPU Throttle Time
0
31
Figure 7-35. Socket Power-Performance Scalability Data
Processor Power-Performance Ratio Data
CPU Socket Power-Performance Ratio
0
31
Figure 7-36. Socket Performance Indication Data
Socket Performance Indication
Socket Performance Indication Data
0
31